Join Professor Ben Glocker, Professor in Machine Learning for Imaging, to discover how causality could help make medical imaging AI safer, fairer and more reliable.
Please register to attend in person. A live stream link for online attendance will be available here shortly.Ìý
We look forward to seeing you on Wednesday 14 October!
51³Ô¹ÏÍø Inaugurals are term-time lectures that celebrate our newest Professors, recognising their academic journey and showcasing their research.
Abstract
Artificial intelligence is transforming medical imaging, promising moreÌýaccurateÌýdiagnosis, earlier detection of disease, and better clinical decision-making. But the path from research prototype to trusted clinical toolÌýremainsÌýdifficult. AI systems can silently fail when new data differs from the data on which they were trained. Changes in patient populations, imaging protocols, and healthcare settings across geographic regions can cause distribution shifts that undermine the reliability, robustness, and fairness of AI predictions.
In this inaugural lecture, ProfessorÌýBen Glocker explores why causality matters for understanding when AI works and when it fails. He will discuss the role of ‘what-if’ reasoning and the use of the latest causal generative models to create realistic counterfactual images that can stress-test AI systems, expose blind spots, and mitigate bias.ÌýUltimately, causalityÌýmay be the missing ingredient that turns medical imaging AI from a promising technology into a life-saving reality.Ìý
Biography
